The number of people affected by natural disasters and the damages that natural disasters cause have increased over recent decades. Climate change threatens to exacerbate this trend. The elderly are likely to be disproportionately affected by disasters, with even greater risk for those with Alzheimer's disease and senile dementia. However, much remains unknown about disasters' effects on adult health and health care costs. For example, no previous study has measured the impact of disasters on health care utilization or expenditures, a potentially important component of the cost of disasters. In addition, existing studies, which overwhelmingly rely on surveys, have been limited by low-frequency data, non-random attrition, small or non-representative samples, errors in recall, and lack of a control group. Overcoming these limitations is important, as public policies and programs aimed at reducing health risks and providing effective assistance following natural disasters depend crucially on understanding the dynamics of the short- and long-run health risks and identifying vulnerable populations. We propose to estimate the short- and long-run effects of Hurricane Katrina on health, health care utilization, and expenditures among the elderly. Our study overcomes the abovementioned limitations by harnessing the power of individual-level administrative Medicare data on the near-universe of elderly (65+) from 1992-2011. Importantly, we can identify individuals living in New Orleans prior to the hurricane, track them regardless of where they move, observe detailed health histories (including health care costs and utilization) before and after the disaster, and construct a control group of unaffected elderly, greatly improving the reliability of our estimates. The completeness of our sample will allow us to look at impacts in potentially vulnerable subpopulations with limited ability to cope with severe disruptions, such as elderly with Alzheimer's disease and related dementias, an important emerging issue for aging research and a high-priority area for NIA. In addition to estimating the health impacts on these subgroups, we will examine the onset and progression of these and other mental and physical health conditions. The project has two Specific Aims: (1) produce comprehensive estimates of the long-run mortality, morbidity, health care utilization, and medical spending impacts of Hurricane Katrina on the elderly and (2) identify the main sub-groups and mechanisms behind the short- and long-run impact of Hurricane Katrina on health, mortality, and health care spending. Achieving these aims can help inform health-improving policy responses.